These results indicate a greater reactivity ... - ACS Publications

hydrolysis of the esters (lO1o-lO1l) than in the previously reported3 acid-catalyzed cleavage of the corresponding methyl ethers (7 x IO6), Id and IId...
5 downloads 0 Views 283KB Size
1548

These results indicate a greater reactivity difference between the tricyclic and anti-7-norbornenyl systems in hydrolysis of the esters (lO1o-lO1l)than in the previously reported3 acid-catalyzed cleavage of the corresponding methyl ethers (7 x IO6), Id and IId, respectively. Since this reactivity difference is presumably associated to a large degree with a relief of ground-state strain in achieving the transition state, it is reasonable to expect (cf. Hammond postulate13) that in the ester hydrolysis the transition state more closely resembles the product carbonium ion than in the acid-catalyzed ether ~ l e a v a g e . ~Hence, one would anticipate a greater relief of strain in attaining the transition state in the tricyclic ester hydrolysis and a greater rate difference when compared to anti-7-norbornenyl. The only two products detected from the hydrolysis of Ib are anti-7-norbornenol and the anti-7-norbornenyl p-methoxybenzoate (IIe) ; however, we have shown that the tricyclic alcohol Ia rearranges readily under the hydrolytic conditions used for the kinetic determinations to afford ~nti-7-norbornenol.~Thus, we carried out the hydrolysis of Ib in 50% aqueous acetone containing 10 equiv of sodium bicarbonate and found a product mixture consisting of 7676 anti-7-norbornenoi and 230,; anti-7norbornenyl p-methoxybenzoate. The nmr spectrum of the product mixture reveals less than 2% of material which could even conceivably be tricyclic in nature. We believe that these data strongly suggest that both Ib and IIe proceed through the same intermediate in the hydrolysis experiments. l 4 Acknowledgment. We thank the donors of the Petroleum Research Fund, administered by the American Chemical Society, for support of this research. (13) G. S. Hammond, J . Am. Chem. SOC.,77, 334 (1955). (14) Professor S . Winstein has independently studied the solvolytic reactivity of endo-tricycl0[3.2.0.0~~~]hept-6-y! p-nitrobenzoate. We thank him for informing us of his results prior to publication.

J. J. Tufariello and R. J. Lorence Department of Chemistry. State University of New York at Bufulo, Buffalo, New York 14214 Received December 4, 1968

The ''c" Route to the 7-Norbornenyl Ion' Sir: One of the most interesting and instructive nonclassical carbonium ions is the anti-7-norbornenyl cation' 11, a bishomocyclopropeny13 ion. This species was originally encountered in solvolysis of bicyclic anti-7-norbornenyl derivatives I (n route).2a,b In recent times, the synthesis of the tricyclic methyl ether2dj4a 111-OCH, has made possible the approach to I1 by the "bo route.2d3e In (1) Research was supported by the National Science Foundation. ( 2 ) (a) S. Winstein, M. Shatavsky, C. Norton, and R. B. Woodward, J . Am. Chem. Soc., 77,4183 (1955); (b) S. Winstein and M. Shatavsky, ibid., 78,592 (1956); (c) S. Winstein, A Lewin, and K. Pande, rbid., 85, 2324 (1963); (d) A. Diaz, M. Brookhart, and S. Winstein, ibid., 88, 3133 (1966); (e) M. Brookhart, A. Diaz, and S. Winstein, ibid., 88, 3135 (1966). (3) S. Winstein, Special Publication N o . 21, The Chemical Society, London, 1967, p 5. (4)(a) H. Tanida, T. Tsuji, and T. I r k , J . Am. Chem. Soc., 88, 864 (1966); (b) H. Tanida and T. Hata, J . Org. Chem., 30, 977 (1965).

FS0,H as a solvent, both routes lead to ion I1 observed directly by nmreZe As regards solvolysis, the available tricyclic derivatives did not permit a solvolytic study under nonacidic conditions which would guarantee observation of the kinetic control product.2d Most recently, the tricyclic alcohol 111-OH has become available,'" and we have now been able to study the hydrolysis of the corresponding p-nitrobenzoate 111-OPNB. The results, which provide increased perspective regarding products, reactivities, and ground- and transition-state free energy relationships in nonclassical solvolysis, are described and discussed in the present communication.

I

111

X

Tricyclic 111-OH, prepared by a modification of the method of T~fariello,~" was converted to its p-nitrobenzoate 111-OPNB by treatment with p-nitrobenzoyl chloride and pyridine in the cold. The III-OPNB,6 mp 121.5-122.5", with an excellent C and H analysis, and displaying the characteristic a-proton quartet at T 4.78 in its nmr spectrum (CCI,), could be isolated when great care was taken to avoid ionizing conditions.' In solvolysis, the tricyclic 111-OPNB is probably the most reactive p-nitrobenzoate ever measured in these laboratories. In 90% acetone at 23", the first-order rate constant for its disappearance ( k , in eq 1) is (6.8 k 0.5) x lo-' set.-' Considerable ion pair return accompanies solvolysis, the observed products being 30% of the relatively inert rearranged bicyclic I-OPNB along with 70% of rearranged bicyclic alcohol I-OH. 111-OPNB

ki

I-OH (70%)

+ I-OPNB (30%)

(1)

The solvolysis of the bicyclic I-OPNB6 is so much slower than that of its tricyclic isomer that we measured its solvolysis rates at 100.0 and 125.1" in 50 and 7076 acetone solvents and then estimated the rate constant at 23" in 907" acetone with the aid of two types of extrapolations. First, the rate constants at 23" in 50 and 70% acetone solvents were estimated by means of temperature extrapolations ; then, the rate constant at 23" in 90% acetone was estimated by means of the Grunwald-Winstein mY relation.8 The results, summarized in Table I, show that tricyclic IIIOPNB is more reactive in solvolysis than its bicyclic isomer by a factor of 8 x lo", nearly 12 powers of 10. (5) (a) J . Tufariello, T. F. Mich, and R. J. Loreace, Chem. Commun., 1202 (1967); (b) J. J. Tufariello and R. J. Lorence, J . Am. Chem. SOC., 91, 1546 (1969). ( 6 ) The melting point observed for 111-OPNB by the ordinary capillary melting point technique, 121.5-122.5', is that of the rearranged bicyclic I-OPNB. The latter, prepared separately from I-OH, had mp 121.5-122.0"and an excellent C and H analysis. Bicyclic I-OPNB, mp 121.8-122.8",was first prepared in these laboratories by C. Ordronneau in 1959. (7) When pure nonpolar solvents and carefully dried base-washed glassware were employed, and recrystallization was from cold petroleum ether, samples of 111-OPNB were obtained which contained > 90% tricyclic ester (the rest being bicyclic I-OPNB) even after the sample was dissolved in CCI4 to determine its nmr spectrum. (8) (a) E. Grunwald and S. Winstein, J . Am. Chem. Soc., 70, 846 (1948); (b) A. H.Fainberg and S. Winstein, 78, 2770 (1956).

Journal of the American Chemical Society 1 91:6 March 12, 1969

1549 Table I.

Summary of Solvolysis Rates

Solvent

50 % acetone 70 % acetone

90% acetone 90% acetone

k , sec-I

Temp, "C Bicyclic I-OPNB

(1.60 k 0.08) x lo-' (1.36 f 0.07) x 5 . 4 x 10-11 (2.82 0.15) x (2.05 0.10) x 10-7 4 . 3 x 10-IZ 8 x 10-14

125.1 1o0.OQ 23. Ob 125.1 100.0' 23. Ob 23.0' Tricyclic 111-OPNB 23.0

(6.8

0.5) x lo-'

"AH* = 28.2 kcal/mol; AS* = - 10.2 eu. *Extrapolated from data a t higher temperatures. cAH* = 30.0 kcal/mol; AS* = - 9.0 eu. dExtrapolated with the aid of the mY relation,s using Y valuessb of 1.398, 0.130, a n d -1.856 for 50, 70, a n d 90% aqueous acetones, respectively.

This makes it ca. loz3 times as reactive as the saturated 7-norbornyl analogZaIV. Also, the tricyclic I11 system is seen to be some 1OI6 times as reactive as its endo-2norbornyl analog' V without the cyclopropane group to provide anchimeric assistance to ionization.

X kre,

X

IV

I

111

V

1

IO"

1023

107

It is instructive to treat the rate comparison between tricyclic 111-OPNB and bicyclic I-OPNB by means of eq 2 and 3, similar to those employed previously for the corresponding methyl ethers.2d In eq 3, K is the equilibrium constant between tricyclic and bicyclic isomers, ( k T / k B )is a reactivity ratio R between tricyclic and bicyclic isomers, and ( k - B / k - T )is a partition factor P representing kinetic control of products of reaction of R+ with X - . kT

111-X

e k-,

R+

+ X-

ke

I-X

(2)

k-0

the same thing in terms of transition-state free energies, the tricyclic transition state lies 3.4 kcal above the bicyclic in the reaction of I1 with CH,OH, whereas it lies 3.5 kcal below the bicyclic transition state in the reaction with p-nitrobenzoate ion. The direction of the difference may be rationalized in terms of an expectedZdlower degree of bond formation at the transition state when I1 reacts with p-nitrobenzoate ion rather than MeOH. The present results illustrate how much flexibility may exist in the ground- and transition-state free energy relationships in reactions of nonclassical carbonium ions, of which I 1 is an example. Regarding hydrolysis products from ion 11, it was shown previously that neutral hydrolysis of bicyclic I-OTs in aqueous acetone leads quantitatively to bicyclic I-OH under conditions which tricyclic 111-OH would surv i ~ e . ' ~ , 'Thus, ~ bicyclic I-OH is the kinetic control product of hydrolysis of I-OTs, disproving the suggestion of Brown''" that the tricyclic alcohol is an intermediate in hydrolysis of I-OTs. The whole product of hydrolysis of tricyclic 111-OPNB in 80% acetone containing excess NaHCO, was now examined by nmr in carbon tetrachloride solvent. The only product besides the 30% rearranged I-OPNB which was detected is the bicyclic I-OH. Using a computer of average transients in the region of thea-hydrogen quartet at z 5.86 to detect tricyclic 111-OH, the 111-OH content was not as large as the threshold value for detection," which we judged to be ca. 0.1%. A control experiment on 111-OH under the conditions for hydrolysis of 111-OPNB and extraction and analysis of the product showed 111-OH to be stable to the whole procedure. Thus, we see that not even the tricyclic system 111 yields tricyclic alcohol as the kinetic control product of neutral hydrolysis. Bicyclic I-OH is essentially quantitatively the kinetic control product of hydrolysis of either the bicyclic or tricyclic system. Thus, Tc-route ionization of I and "oOroute ionization of 111 are shown to lead to the same intermediate even under solvolytic conditions where carbonium ion lifetime is so very short compared to those for nmr observation.2e All the criteria, rates,2 products,2 stereochemistry,zt435a and nrnrZe,',point uniquely to structure I1 for this intermediate.I5 P value for MeO- than for MeOH was commented on previously,2d but the actual value of P for MeO- in MeOH solvent was not avail-

In acid-catalyzed methyl ether equilibration in MeOH, the tricyc1ic:bicyclic reactivity ratio R was observedzd to be 7 x IO6. Since buffered neutral methanolysis of I-OTs yielded 99.7% bicyclic I-OCH, and 0.3% tricyclic 111-OCH,, the partition factor P for reaction of ion I1 with MeOH was 300. This led to an equilibrium constant K between the methyl ethers equal to 2 x 10'. From the present work on the p-nitrobenzoates in 90% acetone, the tricyc1ic:bicyclic reactivity ratio is 8 x IO", some lo5 times as great as the factor in the case of acid-catalyzed ether ionization. Approximating K for the p-nitrobenzoates to be the same as for the methyl ethers, we estimate P for reaction of ion I1 with p-nitrobenzoate ion to be '/400, IO5 times more in favor of tricyclic product than in the case of reaction with neutral methanol." Expressing (9) S. Winstein, et a/.,J. Am. Chem. Soc., 74, 1127 (1952). (IO) Considerably enhanced proportions of tricyclic product are obtained in other reactions of I1 with anionic nucleophiles, such as - B H ~ Z E . I I ~ - CN?b -OMe,Zd and - O C H 2 C 6 H ~ . 5 8 The much lower

able. This is because full control of product formation by MeO- was not attained in the previous work.2d (11) (a) H. C. Brown and H. M. Bell, J . Am. Chem. Soc., 85, 2324 (1963); (b) N. C. Deno, Progr. Phys. O r g . Chem., 2 , 159 (1964). (12) It is interesting that the value of P for reaction of I1 with H?O in aqueous acetone is even larger than it is for reaction of 11 with CHaOH solvent.zd (13) The nmr observations2e914 are especially compelling against the equilibrating tricyclic ion formulation formerly preferred by Brown' l a and DenolIb for ion 11, as well as the closely analogous 7-norbornadienyl cation. For the latter species, recent nmr evidenceI4 on methylsubstituted 7-norbornadienyl cations makes the equilibrating tricyclic ion formulation completely untenable. (14) M. Brookhart, R. K. Lustgarten, and S. Winstein, J . Am. Chem. Soc., 89, 6352 (1967). (15) Tufariello and LorenceSbhave independently studied the solvolysis of 111-p-methoxybenzoate. We thank them for informing us of their results prior to publication. (16) Fellow of the French Centre National de la Recherche Scientifique, on leave from the University of Strasbourg, 1967-1968.

Jean Lhomme,16 Arthur Diaz, S . Winstein Contribution No. 2356, Department of Chemistry University of California, Los Angeles, California 90024 Received Junuarj, 5. 1969 Communications t o ihr Editor